Muscle Tone and Body Weight Predict Uphill Race Time in Amateur Trail Runners

round: Vertical kilometer is an emerging sport where athletes continuously run uphill. The aims of this study were to assess changes in vertical impacts caused by uphill running (UR) and the relation between the anthropometric and lower limb muscular characteristics with speed. Methods: Ten male experienced runners (35 7 years old) participated in this study. In the racetrack (4.2 km long, 565 m high), seven sections were stablished. Mean speed and impact value of sections with similar slope ( 21%) were calculated. The gastrocnemius stiffness (GS) and tone (GT); and the vastus lateralis stiffness (VS) and tone (VT) were assessed before the race. Results: Pearson's correlation showed a linear relationship between vs. and VT (r= 0.829;p= 0.000), GT and GS (r= 0.792;p= 0.001). Mean speed is correlated with weight (r= 0.619;p= 0.024) and GT (r= 0.739;p= 0.004). Multiple linear regressions showed a model with weight and GT as dependent variables of mean speed. Mean impacts decreased signi cantly between sections along the race. Conclusions: The vertical impacts during UR were attenuated during the race. Moreover, body weight and GT were associated with the time-to- nish,

is correlated with weight (r= 0.619;p= 0.024) and GT (r= 0.739;p= 0.004). Multiple linear regressions showed a model with weight and GT as dependent variables of mean speed. Mean impacts decreased signi cantly between sections along the race. Conclusions: The vertical impacts during UR were attenuated during the race. Moreover, body weight and GT were associated with the time-to- nish, which supports that low weight alone could not be enough to be faster, and strength training of plantar exors may be a determinant in UR. Keywords:fatigue; vertical impacts; stiffness; GPS 1. Introduction Trail running or mountain running is a sport that has become very popular in the last years, and the research associated with this long run discipline has also increased [1–4]. One of the modalities of trail running is the vertical kilometer, where the whole track is uphill. In this type of event, athletes must complete an uphill route of 1 km vertical elevation increase. In addition, a minimum average of 20% positive slope must exist and one or more sections of 5% positive slope of the total distance race must be included. Regarding the length and type of terrain, the maximum length must be 5 km and the terrain can vary between different races [5]. Uphill running is a very demanding activity. The athlete must perform positive mechanical power in order to displace their body upward against gravity. This generated mechanical power increases with increasing slope. Compared to level running, in uphill running, due to the lower limb position, hip joint muscles increase their work, whereas the work of knee and ankle remains similar than in level running [6]. The use of elastic energy also changes during uphill running. While most of the energy stored in tendons is recovered in level running [7], running uphill with steeper slopes entails the necessity of raising the center of mass, causing an increase in the positive network generated by the body, considering that the elastic energy stored cannot be used due to the increase in ground contact time. Analyzing the mechanical ef ciency, previous studies have proven than uphill runners show

recovered in level running [7], running uphill with steeper slopes entails the necessity of raising the center of mass, causing an increase in the positive network generated by the body, considering that the elastic energy stored cannot be used due to the increase in ground contact time. Analyzing the mechanical ef ciency, previous studies have proven than uphill runners show around 25% of ef ciency, this value corresponds only to muscle contraction [8,9]. Uphill running cost could explain running performance because this parameter differs from level running. Balducci et al. Int. J. Environ. Res. Public Health2021,18, 2040.

Int. J. Environ. Res. Public Health2021,18, 2040 2 of 9 (2016) [10] studied the in uence of stride length, stride frequency, and body mass index, and no correlation was found between these parameters with the cost of uphill running compared to level running. Nevertheless, in uence of these anthropometric measures was not studied directly in uphill runners. For example, body weight has been shown to be a predictor in the performance of marathon runners [11] and it seems plausible that this factor also in uences the performance of a modality in which it is necessary to raise your body on every step. During running, the foot strikes the ground decelerating the body to zero and gen- erating large ground reaction forces (GRF) [12]. Step by step, impacts are transmitted through the musculoskeletal system. Both passive and active mechanisms act in order to attenuate the shock, minimizing the damage. Although running impacts do not reach extreme values, the quantity of running impacts can be signi cant. Impact peak is very sensitive to leg stiffness and the damping effect of foot (pads) and leg [13].On one hand, the impacts received by the body in each step are attenuated by the bone bending, heel pads, and intervertebral discs as passive dampers; on the other hand, lower limb muscles work actively in order to absorb the impacts. In order to assess running impact acceleration magnitude, several methods could be used, one of them is the tibial acceleration peak, assessed through the placement of an accelerometer in the shank [14], whose value is approximately 8 g [15]. Another method could be accelerometry measurement in the sacrum, presenting smaller values [16]. Through these measurements, the role of active and passive mechanisms to reduce the impact received during running from the shank to the head could be determined [17]. Several studies have analyzed the effect of fatigue on impacts, but their results have not offered a clear conclusion [18]. The possible cause of these discrepancies may be the existence of differences in the way of assessing running impacts between studies. Research articles that have analyzed the ground

during running from the shank to the head could be determined [17]. Several studies have analyzed the effect of fatigue on impacts, but their results have not offered a clear conclusion [18]. The possible cause of these discrepancies may be the existence of differences in the way of assessing running impacts between studies. Research articles that have analyzed the ground reaction forces disagree on the effect that fatigue has on running impacts [19], nding both increases [20] and decreases [21], and offering different explanations for it. The rise could be explained by an increased lower limb stiffness [22]. In addition, another study found a correlation between the pre-activation of gastrocnemius and GRF [23]. The regulation of the lower limb stiffness and the reduced storage of elastic energy could be a possible explanation for the decrease in GRF due to fatigue [24]. Researchers that measured the impacts with accelerometers found an increase in peak impact due to fatigue [25]. The different strategies used to face the fatigue could in uence the responses. In long-distance runners, the global fatigue increases the impacts as well as the local fatigue due to the muscle activity imbalance [26]. Changes in lower limb stiffness can also have an in uence in peak accelerations [14] as well as running technique. Crowell and Davis (2011) [27] found a relationship between running technique and the impacts. The objective of this study was to assess the relationship between anthropometric and the lower limb muscular characteristics with speed, and to analyze the changes in vertical impacts caused by uphill running. 2. Materials and Methods 2.1. Participants Ten male recreationally trained runners participated in the study (age 35 7 years old, mass 68.2 5.3 kg, height 1.77 0.03 m, BMI 21.6 1.4 kg/m 2 ). The inclusion criterion were, at least, one year of experience in trail running races and not suffered from lower limb injuries in the last three months. All the subjects were participants in the uphill race “TurrânSkyrace Pico Las Calmas”, celebrated in Arguis, Huesca (Spain). The participants signed a written consent previous tothe data collection and this study

kg/m 2 ). The inclusion criterion were, at least, one year of experience in trail running races and not suffered from lower limb injuries in the last three months. All the subjects were participants in the uphill race “TurrânSkyrace Pico Las Calmas”, celebrated in Arguis, Huesca (Spain). The participants signed a written consent previous tothe data collection and this study was approved by the Ethics Committee of the University. All procedures followed the Declaration of Helsinki on the use of human subjects.

Int. J. Environ. Res. Public Health2021,18, 2040 3 of 9 2.2. Procedure The uphill race was 4.2 km long and had a 565 m positive slope. In this time-trial race, the runners started to run every 30 s and attempted to complete the racetrack as quickly as possible. In the racetrack, seven sections were distinguished, de ned by six control points (Figure). For each section, the mean speed of the runners was calculated from the time in each control point and the of cial distance track. The GPS device was synchronized to the of cial race start time by means of the detection of the rst change in velocity after a long period of standing.Int. J. Environ. Res. Public Health 2021, 18, 2040 3 of 9 the Ethics Committee of the University. All procedures followed the Declaration of Helsinki on the use of human subjects. 2.2. Procedure The uphill race was 4.2 km long and had a 565 m positive slope. In this time-trial race, the runners started to run every 30 s and attempted to complete the racetrack as quickly as possible. In the racetrack, seven sections were distinguished, defined by six control points (Figure 1). For each section, the mean speed of the runners was calculated from the time in each control point and the official distance track. The GPS device was synchronized to the official race start time by means of the detection of the first change in velocity after a long period of standing. Figure 1. Track of the race with control points (C1–C6), sections (T1–T7), and vertical impact acceleration (VIA_1-VIA_4). The measurements of the skeletal muscle tone and stiffness were taken in a tent provided by the organizer, near the race start zone. 2.3. Measurements Muscle stiffness and tone were collected in lying position (both prone and supine position depending on the muscle evaluated) by a hand-held myometer (Myoton-Pro, Myoton AS, Tallinn, Estonia). Surface Electromyography for the Non-Invasive Assessment Muscle (SENIAM) guidelines [28] were followed to draw on the skin at the testing locations. Medial head of gastrocnemius (MG), lateral head of gastrocnemius (LG), vastus

Muscle stiffness and tone were collected in lying position (both prone and supine position depending on the muscle evaluated) by a hand-held myometer (Myoton-Pro, Myoton AS, Tallinn, Estonia). Surface Electromyography for the Non-Invasive Assessment Muscle (SENIAM) guidelines [28] were followed to draw on the skin at the testing locations. Medial head of gastrocnemius (MG), lateral head of gastrocnemius (LG), vastus lateralis (VL) tone and stiffness were analyzed in the relaxed position. The probe of the Myoton-Pro was placed perpendicular to the skeletal muscle surface in each measurement. Five consecutive measurements were taken at each site, giving the mean stiffness in N/m and the mean tone measured in Hz [29].To ensure validity of the data, a measurement with a coefficient of variation fewer than 3% was accepted, and any measurement above this value was rejected and measured again. The Myoton-Pro offers good to excellent test-retest reliability for lower body tone and stiffness assessment [30]. Muscle tone is calculated using the following formula: í µí°¹= í µí±“í µí±ší µí±Ží µí±¥ (1) Muscle stiffness is calculated using the following formula: í µí±†= í µí±Ží µí±ší µí±Ží µí±¥ × í µí±ší µí±í µí±Ÿí µí±œí µí±í µí±’ âˆ†í µí±™ (2) Figure 1.Track of the race with control points (C1–C6), sections (T1–T7), and vertical impact acceleration (VIA_1-VIA_4). The measurements of the skeletal muscle tone and stiffness were taken in a tent provided by the organizer, near the race start zone. 2.3. Measurements Muscle stiffness and tone were collected in lying position (both prone and supine position depending on the muscle evaluated) by a hand-held myometer (Myoton-Pro, Myoton AS, Tallinn, Estonia). Surface Electromyography for the Non-Invasive Assessment Muscle (SENIAM) guidelines [28] were followed to draw on the skin at the testing locations. Medial head of gastrocnemius (MG), lateral head of gastrocnemius (LG), vastus lateralis (VL) tone and stiffness were analyzed in the relaxed position. The probe of the Myoton- Pro was placed perpendicular to the skeletal muscle surface in each measurement. Five consecutive measurements were taken at each site, giving the mean stiffness in N/m and the mean tone measured in Hz [29].To ensure validity of the

lateral head of gastrocnemius (LG), vastus lateralis (VL) tone and stiffness were analyzed in the relaxed position. The probe of the Myoton- Pro was placed perpendicular to the skeletal muscle surface in each measurement. Five consecutive measurements were taken at each site, giving the mean stiffness in N/m and the mean tone measured in Hz [29].To ensure validity of the data, a measurement with a coef cient of variation fewer than 3% was accepted, and any measurement above this value was rejected and measured again. The Myoton-Pro offers good to excellent test-retest reliability for lower body tone and stiffness assessment [30]. Muscle tone is calculated using the following formula: F=f max (1) Muscle stiffness is calculated using the following formula: S= amax mprobe Dl (2) Raw data were grouped together in order to mitigate the effect of the asymmetries between both limbs (left and right) and to consider possible synergies between muscles

Int. J. Environ. Res. Public Health2021,18, 2040 4 of 9 (the medial and lateral gastrocnemius, the vastus medialis, and lateralis). Four parameters of the participants' muscle mechanical properties emerged out of this procedure: the gastrocnemius muscle stiffness (GS), the gastrocnemius muscle tone (GT), the vastus lateralis stiffness (VS), and the vastus lateralis tone (VT). Body weight was assessed using a Tanita BC-1000 scale. The runners wore a vest during the race with a pocket located in the back (at the height of vertebrae C7) where an Apex GPS device (STATSport Group, Newry, Ireland, UK) was placed. This device includes a 18 Hz GPS and 100 Hz accelerometer in three axes, 100 Hz gyroscope, and 10 Hz magnetometer. The accelerometer signal of the rst 30 s of every race section was evaluated in a custom MAATLAB routine (The Mathworks Inc, Natick, MA, USA) in order to analyze the vertical impact acceleration (VIA) in each slope section. The magnitude of the accelerometer signal was low-pass ltered with a fourth order Butterworth lter with a cut-off frequency of 10 [16]. The mean value of the 15 rst peaks, corresponding to the rst 15 steps was calculated [31]. The mean impacts of sections with similar slope were compared in order to check the fatigue effect changes on the impact magnitude as a marker of fatigue. In this way, Sections 2, 4, 5, and 7 (Table) were analyzed (VIA_1, VIA_2, VIA_3, VIA_4, respectively). Table 1.Slope and horizontal distance of every section of the race besides the type of surface. T1 T2 T3 T4 T5 T6 T7 Distance (m) 1800 800 250 200 220 390 380 Slope (%) 11 22 11 20 21 34 23 Terrain Track Canyon Path Path Path Path Track/path Aside from the impacts, the mean speed (MS) of each section was obtained from the GPS devices' software (Apex Software, STATSport Group, Newry, Ireland, UK). 2.4. Statistical Analysis Statistical analyses were performed using SPSS version 21.0 for Windows (SPSS Inc., Chicago, IL, USA). Descriptive statistics mean, standard deviation (SD), lower 95% con dence limit (LCIL95%) and upper 95% con dence limit (UCIL95%) were

impacts, the mean speed (MS) of each section was obtained from the GPS devices' software (Apex Software, STATSport Group, Newry, Ireland, UK). 2.4. Statistical Analysis Statistical analyses were performed using SPSS version 21.0 for Windows (SPSS Inc., Chicago, IL, USA). Descriptive statistics mean, standard deviation (SD), lower 95% con dence limit (LCIL95%) and upper 95% con dence limit (UCIL95%) were calculated for weight, mean speed, VIA, GS, VS, GT, and VT. Normality of datasets was checked with the Shapiro–Wilk test. Pearson's correlation was calculated and used to determine lineal relationships between all measures. A repeated measures one-way ANOVA were performed to compare VIA along the race (VIA_1, VIA_2, VIA_3, VIA_4) for all participants. The W de Mauchly test was used as sphericity criteria. Regression model (lineal, quadratic or cubic) with the highest order among all those that presented statistical signi cance was considered as the ideal model [32]. Eta square value ( 2) was used for effect size calculation. Bonferroni's post-hoc procedure was applied to locate pair-wise differences [33]. Multiple linear regressions were calculated using a “stepwise” method. Mean speed was considered the dependent variable and weight, VIA, GS, VS, GT, and VT as possible independent variables. Entry and exit criteria were an F probability greater than 0.05 and 0.10, respectively. Residual linearity and independence assumptions were checked with the Durbin–Watson test; values between 1 and 3 in the Durbin–Watson test were consid- ered an acceptable criterion. Homoscedasticity was studied in a standardized residual- standardized prediction plot. Normality of residuals was checked with the Shapiro–Wilk test. Multicollinearity was estimated by a variance in ation factor (VIF), values greater than 10 were considered as excessive multicollinearity. Cases with Cook's distance greater than 1 were indicated as in uential cases and removed fromthe data analysis [34]. All tests were performed with a level of signi cance ofp< 0.05.

Int. J. Environ. Res. Public Health2021,18, 2040 5 of 9 3. Results A summary of the descriptive statistics of the sample can be studied in Table. Table 2. Descriptive statistics of the participants. Standard deviation (SD), lower 95% con dence limit (LCIL95%); upper 95% con dence limit (UCIL95%), vertical impact acceleration (VIA),mean speed (MS), nishing time (min), gastrocnemius stiffness (GS), gastrocnemius tone (GT), vastus lateralis stiffness (VS), vastus lateralis tone (VT). N = 13 Mean SD LCIL 95% UCIL 95% p MS (m/s) 1.81 0.30 1.63 1.98 0.168 Weight (kg) 68.85 4.79 65.95 71.74 0.140 Finishing Time(min) 37.2 6.20 33.45 40.95 0.168 VIA_1 (9.8 m/s 2 ) 2.77 0.34 2.57 2.97 0.570 VIA_2 (9.8 m/s 2 ) 2.34 0.34 2.14 2.54 0.299 VIA_3 (9.8 m/s 2 ) 2.08 0.41 1.83 2.33 0.099 VIA_4 (9.8 m/s 2 ) 1.88 0.32 1.69 2.07 0.099 VT (Hz) 14.80 1.71 13.77 15.84 0.787 GT (Hz) 15.65 1.01 15.04 16.26 0.945 VS (N/m 2 ) 284.87 28.42 267.69 302.04 0.267 GS (N/m 2 ) 280.38 15.09 271.26 289.50 0.211 Pearson's correlation (Table) showed a linear relationship between VT and vs. ( r = 0.829;p= 0.000), and GT and GS (r= 0.792;p= 0.001). Furthermore, mean speed was correlated with weight (r= 0.619;p=0.024) and GT (r= 0.739;p= 0.004). Table 3. Pearson's correlation calculated to determine lineal relationships between all measures. Vertical impact acceleration (VIA), mean speed (MS) (m/s), gastrocnemius stiffness (GS) (N/m 2 ), gastrocnemius tone (GT) (Hz), vastus lateralis stiffness (VS) (N/m 2 ), vastus lateralis tone (VT) (Hz)* p< 0.05;**p< 0.01. Weight VIA_1 VIA_2 VIA_3 VIA_4 GT VT GS VS MS 0.619 * 0.169 0.198 0.200 0.041 0.739 ** 0.006 0.483 0.231 Weight 0.288 0.485 0.370 0.484 0.2790.063 0.030 0.095 VIA_1 0.423 0.1140.256 0.031 0.035 0.2010.162 VIA_2 0.113 0.113 0.106 0.460 0.3540.365 VIA_3 0.544 0.233 0.082 0.248 0.010 VIA_4 0.117 0.4320.021 0.450 GT 0.265 0.792 ** 0.356 VT 0.008 0.829 ** GS 0.104

The study analyzes how muscle tone and body weight affect uphill race performance.